Machine Learning Design Patterns

Valliappa (Lak) Lakshmanan is Google Cloud's Global Head of Data Analytics and AI Solutions. His team uses Google Cloud's data analytics and machine learning products to provide software solutions for business difficulties. He is the creator of Google's Advanced Solutions Lab ML Immersion program.

Sara Robinson is a Machine Learning Developer Advocate in Google's Cloud Platform team. Through demos, online content, and events, she encourages developers and data scientists to use machine learning into their applications.

Michael Munn is a Google ML Solutions Engineer that works with Google Cloud customers to help them create, implement, and deploy machine learning models. At the Advanced Solutions Lab, he also conducts an ML Immersion Program.

Machine Learning Design Patterns's design patterns record best practices and answers to reoccurring machine learning difficulties. The authors, three Google engineers, list tried-and-true approaches for assisting data scientists with typical challenges encountered during the ML process. These design patterns distill the wisdom of hundreds of experts into simple, understandable advice.

You'll find extensive descriptions of 30 patterns for data and problem representation, operationalization, repeatability, reproducibility, adaptability, explainability, and fairness in this book. Each pattern offers a problem description, a number of alternative solutions, and recommendations for selecting the best strategy for your case.

You'll discover how to:

• Identify and prevent common problems encountered while training, evaluating, and deploying machine learning models.
• Represent data for various ML model types, such as embeddings, feature crossings, and others.
• Select the appropriate model type for certain problems.
• Create a strong training loop with checkpoints, a distribution method, and hyperparameter adjustment.
• Deploy scalable machine learning systems that can be retrained and updated to reflect new data.
• Model predictions should be interpreted for stakeholders, and models should treat users properly.

Author: Valliappa Lakshmanan, Sara Robinson and Michael Munn

Link to buy: https://www.amazon.com/Machine-Learning-Design-Patterns-Preparation/dp/1098115783/

Ratings: 4.6 out of 5 stars (from 228 reviews)

Best Sellers Rank: #33,043 in Books

#5 in Business Intelligence Tools

#6 in Machine Theory (Books)

#8 in Mathematical Analysis (Books)